|
I am a loyal customer of Alamo Drafthouse and Cinema - a boutique movie theater chain in the US. One nice perk is they allow you to order with a QR code ordering system from your table. Sadly the search doesn't work that well, when you search for "burger" you get "nachos" as the first result. Well that's an annoyance purpose-built for me. So I took it upon myself to recreate the search, then fix it. Did I build vector search? No. Did I try to improve query understanding? No. What I did instead was learn directly from user traffic. I notice when users searching for 'burger' they click on "Entrees / Sandwiches" (not nachos). Alamo's query distribution likely has a fairly fat head. It means many users type "burger" over and over. We can learn from those interactions. Even for partial queries "bur" we can find the popular, completed query this corresponds to and use that query's data. But, of course, I'm not privy to Alamo's search tracking. Instead I synthesized traffic to my recreated Alamo search system. I added a tracking database on the backend of my app. I spun up a fleet of agents to act as users with specific use-cases (user wanting a beer, another user looking for entrees, etc). It's not LLM as a judge, it's LLM as a user! More details + tricks in the article: https://softwaredoug.com/blog/2026/11/09/fixing-alamo-drafthouse-search -Doug Slack Community · Events · Search consulting · AI-Powered Search training (use code doug-newsletter) You're subscribed to Doug Turnbull's newsletter where I share tips, blog articles, events, and more. You can always manage your profile: |
I share search tips, blog articles, and free events I'm hosting about the search+retreval industry, vector databases, information retrieval and more.
In a world where agents easily drive browsers, why spend time on classic ranking metrics like NDCG? Instead, task an agent with a search use case. See how it fairs in the UI and ranking. In this article, I discuss an alternative eval philosophy. Instead of LLM as a Judge, let agents simulate search users. Observe how they interact with the search app to solve real use-cases. We try to answer the question: can we make search lower effort and easier to accomplish common tasks? Build the...
This week+next there's some upcoming, free search events I'm cohosting: AI-Search Fundamentals (free 'intro to search' training) Today 12PM ET (w/ Me + Trey Grainger) Sign up -- https://maven.com/p/b9e696/ai-search-fundamentals-vector-spaces-matching-ranking This session covers foundational concepts in information retrieval around matching and ranking. We'll cover sparse vs. dense vector spaces, lexical (BM25) vs. semantic (embedding) search, and matching and relevance ranking fundamentals....
I wrote a detailed blog article critiquing the traditional judgment list search evaluation regime. It's impossible to NOT have some kind of sampling bias. IE your labels all come from BM25 or some legacy search algorithm. So solutions can, at best, measure their effectiveness reranking within that sample. Indeed, its hard to think of an alternative. Labeling random results will just end up wasting time (99% of the corpus will be irrelevant). In the end, you need a sampling algorithm to mine...